AB0390 Relationship between body mass index and personality in an early arthritis cohort of patients

Background According to previously reported data from our early arthritis registry (PEARL, Princess Early Arthritis Register Longitudinal), patients with a higher BMI have higher scores in pain and disability scales. We also described some years ago that, in the same population, the structure of personality explored with the PANAS questionnaire (Positive and Negative Affect Scale) affects some outcomes in arthritis measures. Specifically, higher scores on the negative affect subscale associate higher scores in pain and disability scales. Objectives To analyse the relationship between BMI and the structure of the affect in the PEARL cohort. Methods PEARL registry includes patients with early arthritis (less than one year of duration), in whom, sociodemographic, disease related and treatment data are recorded in five protocolised visits. We analysed data from those patients in which the PANAS questionnaire data were available and whose classification, after 2 years of follow-up, were rheumatoid arthritis -according to the 1987 ACR classification criteria – or undifferntiated arthritis. The structure of the affect was evaluated with the PANAS questionnaire administered in one single occasion. This questionnaire is used to evaluate the components of positive (PA) and negative affect (NA). It consists of 20 questions and the score obtained from its administration ranges between 10 and 50 points for each affect. The WHO definition was used for low weight, normal weight, overweight and obesity (BMI Results We analysed the data of the 71 patients for whom PA values were available and of the 65 patients with NA values available. There was not a statistically significant relationship observed between PA and NA and BMI when analysed as a continuous variable (R=-0.12, p=0.25 and R=-0.1, p=0.36, respectively) or as a categorical variable. However, there was a non-significant trend to lower scores for both PA and NA in patients with a higher BMI (figure 1). The multivariate analyses, adjusted for sex, age and study level, also showed no statistically significant relationship between the BMI (analysed as continuous or categorical variable) and the PA and NA; but there was also a non-significant trend that sets a relationship between lower scores of both PA and NA and higher BMI values. Conclusions In our early arthritis registry there is no relationship between the BMI and the structure of the affect evaluated through the PANAS questionnaire. However, it is necessary to evaluate this relationship in a greater number of patients and in different populations before discarding this relationship definitely. Disclosure of Interest None declared